# AaronJackson/vrn

:man:  Code for "Large Pose 3D Face Reconstruction from a Single Image via Direct Volumetric CNN Regression"

Repository: https://github.com/AaronJackson/vrn
Canonical: https://ross.abutalabs.com/products/vrn
Homepage: http://aaronsplace.co.uk/papers/jackson2017recon/
Language: MATLAB
License: MIT
License Family: permissive
Topics: computervision, deeplearning, 3d, reconstruction, face, torch7, computer-vision
Last push: 2022-07-26T00:06:01+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3361, "days_push": 1500, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_readme
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4517, forks 741 (observed 2026-08-28T04:08:52.697038+00:00)

## What it is
Research code for the ICCV 2017 paper 'Large Pose 3D Face Reconstruction from a Single Image via Direct Volumetric CNN Regression'. It uses a CNN to regress a volumetric 3D facial geometry directly from a single 2D image, without needing 3D Morphable Model fitting or dense correspondences.

## Use cases
- reconstruct a 3D face model from a single 2D photo
- generate 3D facial geometry for arbitrary poses and expressions
- research on volumetric CNN regression for faces
- facial landmark localization combined with 3D reconstruction
- reproduce results from the VRN paper

## When to choose
- you need single-image 3D face reconstruction without 3DMM fitting
- you want to study or extend the volumetric CNN regression approach
- you can work with Torch7/Lua and MATLAB research code

## When to avoid
- you need a production-ready, actively maintained 3D face reconstruction pipeline
- you require modern deep learning framework support like PyTorch or TensorFlow
- you need real-time performance on modern hardware

## Facets
- artifact type: library
- maturity: maintenance
- function: computer-vision, machine-learning, deep-learning, image-processing
- domain: computer-vision, deep-learning, machine-learning, artificial-intelligence
- platform: python, lua
- tags: 3d-face-reconstruction, volumetric-regression, cnn, torch7, matlab, research-code, iccv-2017, linux, macos

## Member repositories
- AaronJackson/vrn (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:52.697038+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:20:18.722792+00:00, confidence not recorded.
  - homepage: http://aaronsplace.co.uk/papers/jackson2017recon/ (fetched 2026-08-29T09:06:09.937997+00:00, sha 3e9d33fb1c80)
- Data as of 2026-08-30T08:39:29.467469+00:00.